Public website technology intelligence — one scan, one detailed report.
Website Technology

How to Identify a Website’s Web Server

Learn how to identify a website’s web server using DevTools and curl, spot Nginx, Apache, LiteSpeed, IIS, or Caddy, and interpret CDN headers.

How to Identify a Website’s Web Server

How to Identify a Website’s Web Server

The quickest way to identify a website’s web server is to open your browser’s developer tools, reload the page, and check the Server response header on the main document request. A value such as nginx, Apache, LiteSpeed, Microsoft-IIS, or Caddy gives you a specific lead. If the response says cloudflare, you’ve learned about the network delivering the response. The origin web server behind it may still be unknown. developer.chrome.com

This comes up in real project work more often than you’d think. You might be taking over a WordPress site and need to understand where its redirects are handled. You might be troubleshooting a move to a new host and find that the homepage and the checkout display different server headers. Or you might simply want to know whether an Apache tutorial applies to the project you’re working on. In each case, the answer depends on which URL you check and which layer returned the response.

You can start with the WebBuildFinder website technology detector for a quick overview. For a finding you can document in an audit, verify the response directly using the steps below.

Web Server vs Hosting Provider

A web server is software that handles HTTP requests. It can serve a file, return a redirect, or pass a request to an application. Nginx, Apache HTTP Server, LiteSpeed, Microsoft IIS, and Caddy are examples. A hosting provider supplies the environment where the site runs. That environment could be a shared hosting account, a virtual server, or a managed platform. A CDN or reverse proxy can sit in front of the hosted application and handle the browser’s public connection. MDN

Imagine a WordPress site served by LiteSpeed on a shared hosting plan with Cloudflare in front. WordPress is the CMS, LiteSpeed is the origin web server, the hosting company runs the environment, and Cloudflare handles public delivery. A report that calls all of these the “server” loses the detail you need when deciding where to investigate a redirect, a cache rule, or a missing header.

Check the Server header in Chrome

  1. Open the page you want to check and press F12 or right-click and choose Inspect.
  2. Open Network and reload the page.
  3. Select Doc and click the main document request. Confirm the request URL matches the page you meant to check.
  4. Open Headers → Response Headers and look for server. Also note any CDN or cache headers.
  5. If the page redirects, check the destination document. For a redirect investigation, keep both the redirect response and the final response in your notes.

The request list also contains images, scripts, analytics calls, and other resources. If you select a font file served from a separate hostname, its headers describe that font response. Choose the main HTML document when identifying the server visible for the page itself. developer.chrome.com

Check the same URL with curl

curl -sSL -D - -o /dev/null 'https://example.com/'

Replace the example URL with the public page you want to inspect. The command follows redirects, prints the response headers, and discards the body. It makes a regular GET request. You may see curl -I used for a quick header lookup, but that sends a HEAD request. A regular GET is a better starting point when you want to compare the result with a typical page load. If curl prints several header blocks, use the status lines and Location values to work out which response came from each step.

Nginx

Nginx commonly appears as Server: nginx. Some deployments return a version after the name, while others emit only nginx. Nginx documents that its server_tokens setting controls whether the version appears in the header and on its generated error pages. A bare nginx header is therefore a normal result. nginx.org

The next question is where Nginx sits in the request path. It could serve the HTML or static files itself. It could also act as a reverse proxy and forward some requests to another application or web server. If the homepage advertises Nginx and the HTML contains WordPress theme assets, record both observations. The header identifies a visible delivery layer, and the assets point to the CMS. You’d need configuration or hosting records to know how all requests reach WordPress.

This distinction matters when migrating redirects. Suppose a client hands you a list of rules and says they must live in WordPress because the site uses WordPress. If requests pass through Nginx before reaching the CMS, some redirects could be handled there. Compare the response for an existing old URL with the response for its destination. That can help you identify which responses are involved. Access to the active server and application configuration can confirm which layer owns each rule.

Apache

Apache HTTP Server can advertise itself as Server: Apache or include a longer product string. Apache’s ServerTokens directive controls how much detail the header contains. Its product-only setting returns the name Apache, while other settings can include version or platform details. Two sites can therefore expose very different-looking headers while both using Apache. Apache HTTP Server Version 2.4

If you’re taking over an older website, you may find references to Apache in the handover notes or see an .htaccess file in the project. That file tells you Apache-style configuration exists in the project. The live site may have moved, or another server may handle the public connection. Combine what you see in the repository with the response headers and, if you have access, the hosting configuration.

An old permalink redirect is a good example. If a legacy URL returns 301 with Server: Apache and the destination returns 200 through a different hostname, document both. You can say the redirect response advertises Apache. That’s a narrower and more useful finding than labeling the entire website Apache when you’ve inspected only the old URL.

LiteSpeed

For LiteSpeed, look for Server: LiteSpeed on the main response. You may also see X-LiteSpeed-Cache: hit or X-LiteSpeed-Cache: miss. LiteSpeed’s documentation shows these as expected headers in a configured deployment and notes that the cache header appears when LiteSpeed Cache is enabled for the site. The Server value and the cache status answer different questions about that request. LiteSpeed Documentation

Suppose a WordPress homepage returns Server: LiteSpeed and X-LiteSpeed-Cache: hit. You can document LiteSpeed as the server advertised on that response and record a LiteSpeed cache hit for the page load. If a contact page on the same site shows miss, the server finding can still hold. The cache result changed for that request, which is worth noting if your task involves page speed or troubleshooting stale content.

A WordPress plugin label or an asset path containing litespeed is a weaker clue. It can indicate that LiteSpeed-related software exists in the site’s setup. To understand whether server-side caching is involved in the response you’re analyzing, check the live response headers and repeat the comparison on the relevant public URLs. A site may also send LiteSpeed-related headers through an outer CDN, so match each header to the layer it describes. LiteSpeed Documentation

IIS

Microsoft Internet Information Services can show up as a Server value such as Microsoft-IIS/10.0. You may also come across X-Powered-By: ASP.NET, which Microsoft documents as a header IIS can add. The Server header is the more direct IIS clue. If you’re trying to identify the application framework or programming language, you’ll need evidence from that application as well. learn.microsoft.com

Legacy sites can provide supporting clues through routes ending in .aspx. These can be helpful when you’re looking at an actively used route with matching headers. They deserve less weight when they appear in old links or redirect to a modern dashboard. Sites often keep older URLs available so saved links and search results continue to work after a migration.

Consider a company website with /services/ on the main domain and an account portal on a separate subdomain. You find Microsoft-IIS on the portal while the main domain exposes a CDN header. Report IIS for the portal response. For the public site, record the CDN and leave the origin server open until you find evidence tied to that site.

Caddy

Server: Caddy is a clear Caddy clue on the response where it appears. Caddy can serve static content or reverse-proxy requests to another service. Its documentation also describes ways for administrators to manage response headers. Depending on the deployment, you may see Caddy identified plainly, a backend-generated response with headers modified by the proxy, or no public Caddy label at all. Caddy Documentation

An easy mistake is to infer Caddy from HTTPS or HTTP/3 alone. Multiple web servers and CDN providers support modern web protocols, and the browser can establish its connection with a CDN in front of the origin. Check for a product-specific signal if you want to name Caddy in a report.

If you manage the project and have access to a Caddyfile, check that it belongs to the active deployment and covers the hostname you’re investigating. A configuration file committed for local development can look convincing even though the production site uses a different setup. The public response helps you locate the visible layer. The deployed configuration tells you how that layer is actually used.

CDN Reverse Proxy Complications

A reverse proxy receives the browser’s request and can forward it to an upstream service. A CDN can also return a cached response from its edge network. In either case, the headers reaching the browser may contain information added by the outer layer, passed through from upstream, or a mixture of both. Cloudflare, for example, documents response headers such as CF-Ray and CF-Cache-Status. Those tell you about Cloudflare’s involvement and caching behavior. Caddy Documentation

Public response signalWhat you can reportServer: nginxThe inspected response advertises Nginx.Server: ApacheThe inspected response advertises Apache.Server: LiteSpeed and X-LiteSpeed-Cache: hitThe response advertises LiteSpeed and reports a LiteSpeed cache hit.Server: Microsoft-IIS/10.0The inspected response advertises Microsoft IIS.Server: CaddyThe inspected response advertises Caddy.Server: cloudflare and CF-RayThe response shows Cloudflare in the public delivery path. The origin server is unverified.No Server headerThe inspected response did not disclose a server product in that header.

A cache miss is an especially common source of confusion. A Cloudflare response can report a miss while still going through Cloudflare. The cache status tells you what happened at the edge for that request. If the response still advertises Cloudflare, you have a CDN finding and may need hosting records to identify the origin. Cloudflare Fundamentals docs

Compare only the routes that matter to your work. If you’re investigating checkout, check the checkout document and note any change in hostname. If you’re diagnosing missing caching headers on CSS, examine a real stylesheet request and note who serves it. For a client-owned site, a hosting dashboard, infrastructure configuration, or deployment record can confirm the origin when public responses identify only the edge.

Why Server Headers Are Sometimes Removed

Site operators can reduce the detail in server headers or suppress them. Nginx can omit its version, Apache can show only its product name, and Microsoft documents a setting for suppressing the IIS server header. A reverse proxy can also alter which response headers reach visitors. This means similar public results can come from different configurations, and one deployment can expose different levels of detail across its routes. nginx.org

One reason operators limit server details is to reduce publicly advertised information about the software and its version. A header can also be normalized across several services so responses are consistent. From an audit perspective, a missing header is a concrete observation. It leaves the underlying server unknown unless you have independent evidence. MDN

Write findings so another person can reproduce them. For example, “The public homepage returned Server: cloudflare and CF-Ray. Origin web server unverified.” If the client later confirms that the origin runs LiteSpeed, both facts belong in the report. One describes the public response you observed, and the other describes the origin configuration the client verified.

Start with the exact URL you care about in the WebBuildFinder website technology checker. Then verify the server and CDN headers on that same URL in DevTools. If your audit also covers the application framework, CMS, analytics, or hosting provider, use the complete website tech stack guide to investigate each of those layers.

Need help with your website stack?

Share your current site or a reference build and tell us what you want to create or improve.

Discuss your project